Especially when you pick which tree rings give you the answer you want, and then graft modern data on to it (cough).
And for our ability to measure becoming more precise, I guess that is why they change their methodology regularly (ie HADCRUT2, 3, 4....). We all know good scientists constantly change how they measure things, for consistency. And of course they continually "correct" decades old data as required as well. And if you delete the original data all the better. No going back LOL.
And I work in IT, I know a bit about modelling and algorithms, and though that is far from my expertise, I have to say their ability to model climate is pretty dismal. ALL the models overestimate over the last decades. You would expect in a good ensemble for some to be high and some to be low and reality to be somewhere in between, but they are ALL high. Oh Noes, we're gonna die!!! And it does not take a rocket scientist to figure out why. Models assume as their raison de etre it is all about CO2. Climate is not all about CO2, it's way more complex than that.
I like science as much as anyone, and even ignoring the problems with falsifiability or reproducibility in climate science, few other branches of scientific inquiry leave me as underwhelmed with their predictive ability. If you want to know what it will be like in a century, you have to have some idea of what happens in between. In high school they called it showing your work.
So far it's a fail.